388 SURVEY OF THE COAST 



In choosing a station and in reconnoitring, due attention 

 must be paid to select the most advantageous point of the 

 mountain or place for the view of all the signals required, 

 and the most conspicuous and easily discovered from other 

 places. We must clear, in case of need, the neighbourhood 

 so as to make the signal distinctly visible. Besides this, the 

 ground upon which the instrument is to stand must be very 

 solid. Unless upon rocks, or very dry hills, the solidity will 

 hardly ever be found such as not to affect the level of the 

 instrument by walking near it, or by the different position in 

 which the ol)server stands with respect to it. 



In a simple triangulation, where only few angles are ob- 

 served upon the same point, it may be allowed without much 

 inconvenience to place the instrument at any point near the 

 actual signal which may be favourable for the observations, 

 and to reduce the angles to the signal or station point by the 

 angle of position and distance required. Thus we may use 

 steeples, parts of buildings, or sucli fixed points, (but never 

 trees) as station points. This however must be avoided, 

 when a great number of angles rest upon a point, and it is 

 to serve for detail surveys ; because the consequent calcula- 

 tion of the reductions is actually more tedious than the cal- 

 culations themselves, and because it would be very inconve- 

 nient to keep account of two such near points on the papers 

 intended for the detail surveys. 



It must however be observed, that in general steeples and 

 buildings are always to be avoided if possible, on account of 

 their not giving accurate signals. They should therefore 

 only be used for the determination of their own position, and 

 not to form points in a series of triangles. 



In any case of such a reduction occurring, I used to calcu- 

 late separately all the angles subtended i)y the eccentric dis- 

 tance, at each of the signals upon which angles were oI)serv- 

 ed, and to combine them according to the case of each of tlie 

 angles wliicli 1 had to correct, as the following equation takes 

 place in all cases : — 



